PRODUCT MANAGEMENT
Customer development: a product manager’s top priority
A guide to customer discovery, customer validation, customer creation, and customer building
Google’s graveyard of launched products and many other similar graveyards are full of stories of companies that spent plenty of resources creating a product, or a feature, that customers didn’t want. The most outright example is the late ’90s dot-com boom. However, great products are only created from frequent conversations with potential customers to understand what they really need, a process we will refer to as customer development.
Customer development focuses on
- Understanding customer problems and needs
- Developing a repeatable sales model
- Pivoting the business model to find product market fit or in other words, refining the business model to deliver on customer demand
Customer development aims to help a business quickly find product market fit, identify a compellingly proposed value hypothesis, define the combination of features that would build a caring audience, build a business model that entices customers to buy the offering, and deliver a product that solves a real need at a price point that customers are willing to pay.
Product market fit is the most difficult aspect of creating a company and when you find product market fit, success is almost certain to follow. However, if a company tries to ramp up sales and marketing before finding product market fit, it’ll burn through cash trying to enter a market that doesn’t want its product offering.
The four stages of customer development
The customer development process, is broken into four distinct phases:
- Customer discovery, is to determine who your customers are, and whether the problem you’re solving is important to them. In this phase, you may spend a lot of time conducting primary research, with surveys and interviews, or looking through secondary research. For example, in the case of Uber, Travis Kalanick, decided to build the business model as a private black cab service for himself. Gradually as the service was shared with friends, they began to realize demand from others.
- Customer validation, is when you build a sales process that can be repeated by a sales and marketing team. This process is validated by selling the product to early customers for money. In the case of Uber, customers were paying for the ride from the get go, hence the business model was validated. And for Facebook, in its early days, Mark Zuckerberg was selling banners to local college businesses as a proof that the freemium monetization model will work.
- Customer creation, seeks to increase demand for a product and scale the business. In the case of Uber, the referral bonus program with ride subsidies was the key to its rapid growth, or customer creation.
- Company building, is when a company transitions into a more formalized structure where different specialized departments are created to specialize functions such as sales, marketing, and business development.
The first two phases are the most important parts of the customer discovery process. Getting through these two phases, will guide you to find product/market fit, where most start-ups and new products fail to reach. Each of these phases assumes that it may take several attempts to get it right.
I — Finding problems
Sources of idea
Great product ideas often come from a painful problem someone is experiencing. On day one, Dropbox, Uber, or Airbnb were scratching the founders’ own itch, and as things progressed, they came to realize that there were a lot of other people with the similar itch out there.
Product ideas can come from both internal and external channels to your company such as the sales team, customer service, data analytics, engineering team, or customers. For more established companies, ideas may be cultivated from industry trends and competitors actions. For example, following Snapchat’s success, Facebook, Instagram, Whatsapp, and YouTube built new features that imitated Snapchat’s expiring photo functionality.
A product manager’s role is to take a critical look at any product idea and determine whether it’s a real opportunity for the company, analyzing the potential opportunity and strategic fit with the organization’s capabilities.
In a startup setting, the founders tend to take on the product manager role
Hypothesis-driven process
Most products fail because an aspirational entrepreneur or product manager thinks that he/she has identified a problem, is consequently struck with a ‘great solution idea’, and assumes that tons of other people must need this solution too. But to turn this into a reality, the product manager needs to figure out whether it is a real problem or not and whether the solution does actually solve that problem. The goal of customer discovery is to figure out:
- Who the customers of a product are
- Whether the solution is important to the customers
- Make sure that customers are willing to pay for the offering
When a product manager wants to discover customers, he/she needs to take on a hypothesis-driven approach in solving the problem, breaking the process down into four distinct sections:
- Stating a problem hypothesis
- Testing the problem hypothesis
- Offering and testing the solution
- Verifying the hypothesis or pivoting
Where many products fail is in validating assumptions. Product managers are constantly making assumptions about who the users are, what their problems are, and why they’d want to pay for the service. To prevent leaps of faith, we need to create hypotheses around assumptions, go out and test those hypotheses with real people, and get real answers as to whether or not we are on to something.
Customer discovery is basically applying the scientific method of ‘Validated Learning’ or ‘hypothesis-driven problem solving’ to business challenges. Through customer discovery, product managers aim to generate measurable data to tell whether or not the hypothesis is right or wrong.
A business hypothesis is your assumptions with more details including:
- Who you think the users of your product are?
- What problems customer are facing that you intend to solve?
- How your product actually solves the problems?
- Why customers might be willing to pay?
So, first talk to your users to understand their problems and then begin testing whether or not your idea or product solves those problems. The most important thing is that you should not be trying to convince people to use your product, but be concerned with learning as much as you can about these people and their problems. If you feel confident that you’ve identified real problems, you can begin testing your ideas as solutions.
If at any point in the process, you cannot confirm that your assumptions were correct, then you simply go back to the beginning and try again with a different assumption and hypothesis. By constantly running through this lean process of customer discovery, you can quickly find the best way to solve real problems for real users before building anything.
The hypothesis user story
Before testing any idea a product manager needs to develop a problem hypothesis, not a product hypothesis, by talking to potential customers and figuring out their needs. For instance, Evan Williams, the founder of Blogger, saw that while blogging was becoming more popular, most people weren’t interested in creating long form content, so he assumed that people would be interested in a micro-blog platform with 140 characters, and that’s how he ended up creating Twitter. ‘An observation to an assumption, about a problem hypothesis, and a starting point for customer development.’
To create a problem hypothesis, use this model:
As a [user type] I want [behavior] so that [outcome or benefit]
For “As a [user type]” part:
- You want to be as specific as possible in defining the role of the user.
For “I want [behavior]” part ask these questions:
- What does your user need or want to be able to do?
- What behavior are they willing and hoping to be able to show?
- What is the core desire driving them to evaluate products, including yours?
For “so that [outcome or benefit]” part ask yourself:
- What is the ultimate outcome or benefit to the user? This is generally the real need they are looking to solve. Will it save them in costs or time? Will it make them more money? Will it give them status? Improve their role?
This format helps create a hypothesis in the form of a user story. The user story is written from the standpoint of a user. Let’s use the Twitter example to write a product hypothesis in the form of a user story.
‘As a socially active blogger, I want to be able to share brief ideas so that my voice is heard by as many people as possible, without spending a ton of time writing.’
By developing a problem hypothesis in the form of a user story, it begins to provide insight into the product you can create to solve the user’s problem. Once you begin talking to real people who fit the user type in your problem hypothesis, you can write survey and interview questions to gain insight into their needs. This way, you’ll be able to validate whether or not your assumptions are correct and if you’re on the right track to defining a product worth building.
II — Conducting customer research
In startups and established corporations, user research is the most frequently overlooked aspect of creating a product. Without performing user research, you’re betting on your ability to totally guess what a user needs and whether people will flock to it — and that rarely happens. Customer research can come from both secondary and primary sources of information.
Primary sources include surveys and customer interviews, in individual or group formats. Primary sources of research are key to product development and product managers need to gather information directly from customers if they intend to make successful business decisions. Secondary sources include anything that you don’t learn directly from customers, so this may include:
- Academic research
- Private research firms
- Consumer reports
- Interviews with experts
- Industry trends reports
- Competitor’s moves
Before jumping in and doing all the hard work yourself, it’s wise to see if someone has provided a head start. They may be able to help you avoid pitfalls, refine your hypothesis, or further define your target user based on psychographics.
Once you’ve determined that you’re on the right track with secondary research, you can start doing primary research. This is where you talk to users directly and where the bulk of your time should be spent. There’s just no replacement for direct feedback from users. They know their needs better than anyone else and will usually gladly talk to you about them. If you’re worried that talking to potential customers will reveal bad news about your idea, then it’s probably not an idea worth pursuing anyway.
Be targeted
In reality, your product won’t be for everyone, so it’s important you know who it is for. If you use feedback from people who aren’t your target users, you may create the wrong product. Try to define your target customers in well refined and specific segments that include demographics, psychographics, attitudes, habits, and needs. For example:
Women, age 25–35, educated, housewives, with +2 children, household income of +100k annually, located in upper class segments on the town, social activists, reads books of a regular weekly basis, environmentally conscience, tech-savvy but in touch with traditional family bonds, and in needs to building communal bonds in her vicinity with other housewives.
This is quite granular and specific and can be a great starting point. When your idea starts to scale, you can expand your target audience. For example, in the case of Airbnb, you may say that their target users are people who saw view unused space as an asset that could be leveraged to earn an extra income and it doesn’t matter whether someone is 20 years old or 60 years old, as long as they recognize that their unused space could make them money and believe that renting to strangers is safe, you are expanding the target market.
Notes to consider when conducting primary research
Care needs to be exercised when creating surveys or conducting interviews. If questions are biased, leading, or skewed, you simply won’t get responses that provide meaningful insights.
Surveys are great for measuring attitudes and collecting quantitative feedback, but they are not good for understanding the underlying drivers of customers’ problems or needs.
On the other hand, asking someone the open-ended question, ‘please describe your current job satisfaction level’, in a survey will not be useful. Few people will bother to answer. These types of open-ended questions are best left for interviews, where you allow the user to speak freely right in front of you. This approach leads to more insights because you can tailor the conversation based on their responses.
Before running surveys or interviews, make sure that you run an online screener, using SurveyMonkey for example, in order to target the exact people you want to talk to . Try to:
- Create a list of the characteristics of people you want and don’t want to talk to.
- Create specific criteria that matches these characteristics of the people you want to interview.
- Come up with questions that help you decide whether someone meets the criteria or not and place them in your screener
Good customer interviews help shed light on who your customers are and whether or not the problem your product solves is important to them.
III — Primary research — surveys
Components of a good survey
Surveys are great for quantifiable information and because they can be made quickly and cheaply, product managers use them too often to collect customer information, bearing the risk of running an improperly designed and used survey that could lead to misleading or completely worthless results. Surveys are great for three things:
- Measuring attitude, intent, or task success. For example, how does the user feel about this? How important is this to the user? Or how successful is the user at completing this task?
- Tracking changes over time. For example, what were the user’s attitudes before the feature launch? Or what were the user’s attitudes after the feature launch?
- Quantifying problems users are experiencing. For example, users have reported an issue with this part of the software, but how many users are experiencing that issue?
Surveys aren’t good for:
- Discovering the reasons behind what your users care about and what they really need
- Understanding whether someone can use your product effectively
- Understanding actual user behavior with your product
Problem and product discovery are better suited for user interviews
Until you have a product in the market, even an early version, you should spend the bulk of your time in user interviews. If you interview potential customers and discover that several of them have the same issues, you can incorporate that insight into a survey to get a larger sample.
Good surveys are:
- Short and to the point
- Focused on specific insights and purposes
- Grouped, with similar questions placed together
Start broad and gradually ask specific questions
Some common bias creating pitfalls when creating surveys include:
- Asking leading questions. Questions like, would you use a product like this, lead to biased responses and make your data potentially inaccurate.
- Asking whether they agree with a statement or not. Instead of asking a user if they agree with a statement, ask them to rate how they feel about a subject. Here’s an example. Ask, ‘how satisfied are you with your current cable company’.
- Asking double-barreled questions, meaning two questions in one sentence, such as, do you have cable and Internet at home?
- Asking open-ended why questions, such as, why do you watch cable? Open-ended questions should be left for user interviews.
Creating a survey
One of the most frequently used surveys is the Net Promoter Score, or NPS, survey. The NPS survey seeks to measure customer experience and predict your business’ growth. It’s often just one simple question:
How likely is it that you would recommend this product to a friend or colleague?
The user is given a 0 to 10 rating scale between not at all likely, to extremely likely. This is a perfect example of a good survey question. It uses a rating scale, avoids bias, and gives the survey owner quantifiable data to measure and understand.
Analyzing responses to a survey
Do not study the average response for each question, it’s usually meaningless. Dig deeper into assessing the distribution of responses by assessing responses using the following guideline:
The significance of a response, or the score to a question is:
- Note (1): This method is mainly used for assessment of the NPS, but you can apply it to any other question in order to get a better assessment of the distribution of responses in your answers.
- Note (2): Make sure that you use the correct sample size to consider the statistical significance of your survey.
IV — Primary research — interviews
Components of a good interview
- Interviews are meant to be mostly free-flowing and fun and not boring
- Strive for a comfortable and enjoyable conversation
- Keep conversations as one-sided as possible; the more time you can keep them talking while you remain silent, the better. If you’re talking, you are not learning
- Build rapport with the user
- Ask open-ended and non-leading questions. These are your classic who, what, when, where, why, and how questions. People like to talk about themselves. So, by asking open-ended questions and not letting them off the hook with a simple yes or no
- Start with some easy questions that aren’t too personal or difficult and as the user becomes more comfortable, you can ask more direct or intense questions, and when you feel good about the information you’ve received, gradually come back to simpler questions so you don’t end the conversation abruptly.
- Don’t feel confined to stick with your prepared questions, be flexible and move with the direction the interviewee is taking the interview and dig deep into what you are looking for.
- If they cannot make a statement for themselves, it’s not an accurate concrete piece of information. If the user is responding with, some people think, or want, or say, be sure to ask them for an example for what they’ve experienced or thought personally.
- Do not create bias. Do not pitch the user on thinking a certain way. You aren’t conducting this interview to sell your product at this point.
The question template
By creating a question template, you make sure that you start each interview intentionally to set yourself up for success. This keeps you organized and makes it easier to summarize your findings later. A question template is typically structured as follows:
- A list of learning objectives for the interview. What are you trying to learn by talking to this person?
- An introduction that lays the land. Before asking any questions, you should start with a basic introduction to share expectations and lay some ground rules.
- Starts with low intensity warm up questions and gradually intensifies the difficulty level of questions.
- Asks for details of the interviewee in case of a need for follow up or other round of interviews.
Conducting interviews
- A successful interview is the result of a combination of good questions and the way you conduct yourself as the interviewer.
- Your body language and presence in the room needs be friendly, engaging, and casual.
- Take thorough notes or record the interview on audio or video so that you stay focused on the individual and not just your own note taking.
- Stay neutral so you don’t bias responses and encourage interviewees to keep talking. This can be as simple as leaning forward in your chair and nodding your head. By showing your interest, you’re more likely to encourage the interviewee to continue expressing themselves freely.
- Watch the body language of the person you’re interviewing. Do they seem open and comfortable? Or are they facing away from you and avoiding eye contact? If so, it’s important to slow down and regain comfort in the conversation.
- Make eye contact, not too intensely, and along the way, express gratitude for their input and encourage them to further elaborate.
Mining for patterns in results
The most important part of analyzing interview results is looking for patterns. You have to look for similar evidences in responses. There are four specific patterns you’re looking for. You’re looking to find out whether your interviewees:
- Mention the same problem
- Indicate they’re already actively seeking solutions
- Say they’re unhappy with solutions they’re currently finding
- Have a budget already available to solve the problem
Based on the patterns identified above, you will have five varying categories of customers:
- Those who are not mentioning your problem hypothesis at all
- Those who are mentioning that they frequently experience the problem you assumed
- Those already looking for solutions to the problem
- Those who are unhappy with the solutions available to them
- Those who already have a budget available to solve this problem
Now you have to segment your customer based on their responses, measure the severity of the problem for each segment, and identify the opportunity. You want to avoid jumping into conclusions about patterns until you’ve totally finished all of your interviews. As you conduct more interviews, only gradually will you witness patterns emerge. These patterns may or may not validate your problem hypothesis, but they’re the key to zooming in on a real opportunity that solves problems for your users.
V — To proceed, pivot, or stop efforts
Once you’ve talked to customer, you can finally confirm whether or not your problem hypothesis was correct, which will fall into one of these three categories:
- You move on aggressively: Your assumptions were correct and were validated with real users
- You’ll need to make adjustments with minor pivots pivot: Your assumptions were close, but with some key points of difference that need to be reassessed
- Stop investing more efforts or make a major pivots: Your assumptions were totally off and users didn’t mention the assumptions as problems at all.
By discovering that your assumptions are wrong at this stage of the process, you’ve saved your company a ton of resources and can take another shot at solving a different problem. It’s rare that your assumptions are totally correct, however, by repeating this cycle until you discover a problem that people really want solved, you’ll stay lean and give your company the best shot at creating a product that people desire.
Product graveyards are full of stories of companies that spent plenty of resources creating a product, or a feature, that customers didn’t want. The most outright example is the late ’90s dot-com boom. However, great products are only created from frequent conversations with potential customers to understand what they really need, a process we will refer to as customer development.
